[Pw_forum] drho and D3 calculations

[CLAUDIA BUNGARO]

Dear Jonathan,

one problem could be  that the namelist variables controlling
input/output file names have been  changed in the version 1.2.0
for pw and ph. The variable 'filpun' has been removed and the
variable 'prefix' has been introduced.

It seems that in the namelist for d3 the 'filpun' variable still exist
and the variable 'prefix' does not exist.

Maybe the d3 code has not been modified to reflect the changes in
pw and ph and it creates problems?

This may explain the error you get running d3,
which says it cannot read the file that saveall.f90 writes and reads.


Claudia

__________________________________________________
Dr. Claudia Bungaro
Dept. of Physics and Astronomy, Rutgers University
136 Frelinghuysen Road, Piscataway,  NJ 08854-8019

tel: +1 732 445 4197          fax: +1 732 445 4343
email: bungaro at physics.rutgers.edu
__________________________________________________


On Fri, 6 Jun 2003, Jonathan Breeze wrote:

> Dear PWscf users,
>
> I have been trying for a long time to calculate the 3rd order anharmonic
> coefficients at the Gamma point.
> For a simple system like MgO using the following file for pw.x:
>
> Mgo
> Magnesium Oxide
>  &control
>     calculation='scf'
>     restart_mode='from_scratch',
>     prefix='mgo',
>     pseudo_dir = '/usr/people/breezejd/pwscf/pwscf-parallel/pots/',
>     outdir='/usr/people/breezejd/pwscf/pwscf-parallel/tmp/'
>  /
>  &system
>     ibrav=  2, celldm(1) =7.27303, nat=  2, ntyp= 2,
>     ecutwfc =44.0, occupations ='fixed',
>  /
>  &electrons
>     conv_thr =  1.0d-8
>     mixing_beta = 0.7
>  /
> ATOMIC_SPECIES
>  Mg  24.3050 Mg.pz-vbc.UPF
>  O   15.9994 O.pz-mt.UPF
> ATOMIC_POSITIONS
>  Mg 0.00 0.00 0.00
>  O  0.50 0.50 0.50
> K_POINTS{automatic}
>  4,4,4,0,0,0
>
> Everything goes well after the SCF calculation. I then run the phonon
> code at the Gamma point to produce a drho file.
>
> phonons of MgO at Gamma
>  &inputph
>   iverbosity=1,
>   prefix='mgo',
>   amass(1)=24.305,
>   amass(2)=15.9994,
>   tmp_dir='/usr/people/breezejd/pwscf/pwscf-parallel/tmp/',
>   fildrho='mgo.drhoG',
>   fildyn='mgo.dynG',
>  /
> 0.0 0.0 0.0
>
> Phonon produces the dynamical matrix which seems ok and four drho files:
> mgo.drho, mgo.drhoG1, mgo.drhoG2, mgo.drhoG3.
>
> mgo.drho look like this: which seems a little truncated.
>
>       24      24      24      24      24      24       2       2
>      2  7.27303000  0.00000000  0.00000000  0.00000000  0.00000000
> 0.00000000
>       235.8216583131        4.0000000000       44.0000000000     0
>     1      Mg     2.00     0.0000000000  0.0000000000  0.0000000000
>     2      O      6.00     0.5000000000  0.5000000000  0.5000000000
>  13824*0.E+0
>
>
> The  other three are quite large (about 210K).  I have no idea what the
> numbers in these files mean.
>
> Anyway, I have tried to run D3, the input file for this is:
>
> third order derivatives of dynamical matrix for MgO at Gamma
>  &inputph
>   iverbosity=1,
>   tmp_dir='/usr/people/breezejd/pwscf/pwscf-parallel/tmp/',
>   filpun='mgo.pun',
>   fildyn='mgo.d3dyn',
>   fildrho='mgo.drhoG',
>   fild0rho='mgo.drhoG',
>   amass(1)=24.3050,
>   amass(2)=15.9994,
>  /
> 0.0 0.0 0.0
>
> The cards load in OK but then I get an error;
>
>
>      Program D3TOTEN1.2.0 starts ...
>      Today is  6Jun2003 at  7:47:46
>
>      Parallel version (MPI)
>      Number of processors in use:      1
>
>      UltraSoft (Vanderbilt) Pseudopotentials
>
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>      from saveall : error #        21
>      reading file
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>      stopping ...
>
> I'm sorry for this long email, but I have tried posting on the forum a
> number of times and I have never received a single response.
> I am running short of time now on the project and may have to abandon
> using the D3 (2n+1) method in favour of another method such as frozen
> phonon.
>
> Please help!!!
>
> Yours pleadingly,
>
> Jonathan Breeze
>
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>